Sometimes you want to listen to two different signals, on two independent receivers, from a common antenna. To do that, you need an RF signal splitter, as shown below. In this issue, I explain the splitters that I built.

When the signals received at the antenna are applied to the two-output splitter (hereafter 2-splitter), the power of the output signals are reduced by 3 dB (half), even though the splitter circuit itself has no loss. Common circuits using an RF transformer are shown below.   

See circuit diagram (a) above. The impedance at the center tap position on T2 is 25Ω when the input impedance is 50Ω. In order to match the impedance between the center tap on T2 and T1 I needed to build a transformer with the impedance of 50Ω/25Ω. The transformer turn ratio is
√2 (1.41):1. A resistor connected at the output end works to keep the two output levels of (b) and (c) constant. A three-output (3-splitter) circuit is shown above, and works similar to the 2-splitter. I tried to build a 2-splitter and a 3-splitter. A 2-splitter needs two transformers, and 3-splitter needs four transformers. I got ferrite cores to build the transformers from a shop on the internet. The sizes of the ferrite cores are 14mm x 23mm, and they are two-hole types. See the picture below of the ferrite cores.

For the 2-splitter, T2 is a transformer with a center tap. The turn ratio is 1:1. T1 has a turn ratio of
√2(1.41):1. 

3-splitter unit and its transformers (two-hole types)

After I built the 2-splitter, I measured the frequency characteristics for the circuit loss against the frequency.

As you can see, the characteristics are disappointing and not what I expected, especially the level of loss. If the loss were 10 dB or less, I intended to accept it, but the result is more than 10 dB over the entire frequency range of 1 GHz. I used the right two-hole type cores for RF. I tried several times to make the coils, and resoldered the coils to the PC board several times, but I could not see a big improvement. The problem may have been the material of the two-hole type cores, or the way I wound the coils.

I took out T2 and tried to measure the characteristics of only the transformer, using a tracking generator. Originally T2 should have little loss, but there was a loss of more than a few dB. I could not improve it, even by trying various ways of winding the wire, and so on.

I found that it cannot be improved unless I review basic parts such as core material and material of windings.

Since I was unsuccessful with the transformer type splitter, next I experimented with resistor type circuits. Circuit diagrams for the resistor type splitter are shown below.

The output of the 2-splitter has a 6 dB loss, and the output of the 3-splitter has an 8.5 dB loss against the input levels. At -6 dB, the S-meter indication of the signal strength was down by 1 unit, and at -8.5 dB it was down approximately 1.5 units. At this level, I used the splitter unless it is a very weak signal. It seems that weak signals will not be heard with that loss.

For the S-meter reading versus signal level, see the chart at the end of this article.

In this circuit, using chip type resistors, the frequency characteristics are not too bad. Especially in the low frequency range, the characteristics are good compared to using ferrite core transformers. For a 2-splitter, this method may be good if it does not have much impact on receive sensitivity, even if the S-meter readout drops by about one unit.

The frequency characteristics of a resistor type 2-splitter are as follows.

The graph on the left shows the characteristics from 1 MHz to 1 GHz. It seems that there is some disturbance below 200 MHz, so that part of the graph is enlarged on the right.

I do not understand why it is disturbed, but it seems that there is not much effect on overall performance.

The following figure shows the SWR characteristics when the resistor type 2-splitter is used.